Page 34 - Summer2017

Insect Bioacoustics
spectra show that the frequency of spontaneous oscillations matches the dominant frequency of the songs of the species (Riabinina et al., 2011). The passive mechanical properties of tree cricket wings are such that the song frequency varies considerably with temperature (Mhatre et al., 2012). Active mechanical processes in their ears result in a parallel shift in auditory tuning, allowing females to remain selectively sen- sitive to male songs over the wide temperature range that these insects experience (Mhatre et al., 2016).
Sound Localization
Most insects need to determine sound location in three di- mensions: azimuth, elevation, and distance. It is generally assumed that distance estimation is based largely on the per- ceived stimulus amplitude, which decreases with distance from the source, although distance-dependent change in the signal spectrum caused by frequency-dependent attenuation is also a possibility (but only for those insects capable of fine spectral analysis, such as katydids). Although there is con- siderable behavioral evidence that at least some insects can determine the elevation of a sound source, little is known about the underlying mechanisms. In particular, insects lack the elaborate outer ear structures of mammals (pinnae) and birds (facial ruffs of owls) that generate elevation-dependent spectral cues (Roffler and Butler, 1968; and as noted above, only some insects could utilize these cues). One possibility is that they use behavioral strategies, such as twisting their bodies so as to generate left-right differences in orientation toward an elevated or depressed sound source. Another pos- sibility for flying insects is that the flapping wings might dif- ferentially affect acoustic input to the ear for elevated versus depressed sound sources (Payne et al., 1966).
The mechanisms underlying determination of sound-source azimuth have received much more attention. As mentioned earlier, azimuth-dependent ITDs are likely to be too small to be resolved by insect nervous systems. In some cases, an insect’s body may be sufficiently large relative to the sound wavelength to generate IIDs of at least a few decibels. Such is the case, for example, for the ultrasound (i.e., short wavelength) avoidance responses of large moths and similarly sized insects.
In other cases, the discrepancy between body size and wave- length is so large as to make sound localization based solely on diffraction unlikely. The two best studied examples of this are crickets and the parasitoid flies for which crickets are hosts. The cricket songs to which both insects orient have a wavelength of about 7 cm compared with an interaural dis- tance in crickets of about 1 cm and of only 500 μm in the fly.
Figure 5. The cricket pressure-gradient system in situ (top) and sche- matically (bottom). Vibration of the ipsilateral tympanum (IT; the eardrum on the same side as the sound source) is affected by sound acting on its external surface and by sound reaching its internal sur- face via tracheal paths from the ipsilateral (IS) and contralateral (CS) spiracles. CT, contralateral tympanum; CM, septum that joins the left and right tracheal branches. Dashed lines indicate that the acoustic tracheae have been truncated. Modified from Michelsen (1998).
Nevertheless, both insect groups are capable of exquisitely fine determination of sound azimuth as a result of coupling between the left and right ears.
Pressure Gradient Ears of Crickets
As in katydids, the cricket respiratory system is adapted for auditory function where a specialized acoustic trachea links a spiracle to the inner surface of the eardrum. Here, however, the main result is improved directionality rather than increased sensitivity. Vibration of each eardrum is driven mainly by sound that reaches it along three routes: directly to its exterior surface, indirectly to its interior sur- face through the tracheal route on the same side, and via the acoustic spiracle and trachea on the opposite side. The latter route is possible because the left and right acoustic tracheae abut at the midline where they are separated by a thin sep- tum (Figure 5).
The relative amplitudes and phases of the three inputs vary with the sound direction because of differences in arrival times at the three loci as well as the highly frequency-de- pendent transmission characteristics of the tracheal path- ways (Michelsen, 1998). For cricket song frequency, which is about 5 kHz, summing the three inputs results in direction- dependent variation of vibration amplitude approaching 20 dB, far greater than the maximum of 1-2 dB possible from diffraction alone.
32 | Acoustics Today | Summer 2017